Corpus COFLA

Corpus COFLA是一个专门为计算研究弗拉门戈音乐而创建的数据集，由庞培法布拉大学和塞维利亚大学的研究人员共同开发。该数据集包含超过1800条音频记录和相应的编辑元数据，总计约95小时音乐，涵盖了20世纪所有公认的经典弗拉门戈录音。数据集旨在为特定音乐信息检索任务的开发和评估提供注释子集的池。Corpus COFLA不仅支持计算研究，也适用于传统音乐学或语义分析，旨在通过自动描述和分类弗拉门戈音乐，促进音乐数据库的自动化索引和该流派的传播。

Corpus COFLA is a dataset specifically created for computational research on flamenco music, jointly developed by researchers from Pompeu Fabra University and the University of Seville. It comprises over 1,800 audio recordings alongside corresponding curated metadata, with a total duration of approximately 95 hours, covering all recognized classic flamenco recordings from the 20th century. The dataset aims to provide a pool of annotated subsets for the development and evaluation of targeted music information retrieval tasks. Corpus COFLA not only supports computational research but also applies to traditional musicology and semantic analysis, and is designed to promote automated indexing of music databases and the dissemination of this musical genre through automatic description and classification of flamenco music.